TODO:

• Add papers trying to replicate
• Binderize
• Add instructions on how the notebooks/scripts work and what order they should be executed in
• Add some scientific background/context
• Discuss issues with reproducibility if they can't be fixed

TODO:

• Build a Circle config file
• Add suitable tests

Figures that are not Reproducible

1. Fig. 2 does not have the correct traces compared to paper.
2. Fig. 3: Order of magnitude issues for Piro & Ott solution and Bucciantini solution is failing.

Probable causes

Version issue between python versions 2.7.10 and 2.7.15 (and related numpy, scipy and matplotlib libraries). When reverting to the lower version, binder cannot install the required downgraded versions and scipy and numpy. Therefore, these graphs are currently not reproducible.

Add new notebooks to create figures from Gibson et al. (2017)

• Fig. 1: Add markdown discussion
• Fig. 2: Disc mass/omega/fastness parameter vs. time
• Fig. 3: Dipole torque comparison
• Fig. 4: Type identification
• Fig. 5: Type comparison with Gompertz et al. (2014)

Problem

Attempting to follow this binder example so that the files contained within the folder /magnetar will be launched as a package and are available to the notebooks and scripts.